Hitherto, in a data base system of a computer, there is widely used a relational database (RDB). Recently, however, with the advance of the associated technology of the Internet and multimedia, there is increased such a pattern that the relational database (RDB) is changed over to an object database (ODB) suitable for those processing.
FIG. 22 is a view showing an example of RDB data.
An RDB comprises, usually, an assembly of data files of two-dimensional table format, which are referred to as “tables”.
Part (a) of FIG. 22 shows a part of tables of an RDB addressed as “makers”. Part (b) of FIG. 22 shows a part of tables of an RDB addressed as “beverages”. Columns of “maker codes”, “maker names”, . . . , and “annual sales” displayed on the uppermost stage of the table referred to as the “makers” shown in part (a) of FIG. 22 are addressed as items. “001”, “Karyn”, . . . , and “1200 hundred million” appearing on the first line are data associated with the respective items mentioned above. A set of those data is referred to as a record. A single table “makers” is constituted of a plurality of records as mentioned above.
Part (b) of FIG. 22 shows a part of tables of an RDB addressed as “beverages”. The “beverages” table has 12 items such as “names”, “large classification”, . . . , and “country codes”. Of those items, seventh item is named “maker codes” which is the same name as the first item of the “makers” table. In this manner, the “makers” table and the “beverages” table, which have the same item “maker codes”, are associated with each other by means of comparing the values of the “maker codes” with one another, for example, “001” of the first record of the “makers” table with “001” of the first record and the fourth record of the “beverages” table, so that a relational database system is constructed.
Thus, in the RDB, an item name and the associated data type and number of characters are defined for each table, and the definition information is described in a so-called RDB repository (definition information management list). In the RDB, usually, operations such as registration, supplement, deletion, alteration, retrieval of data are performed in accordance with the database operation language referred to as the SQL. For a regular form of data processing work involving a database operation, an application program written by the SQL is created and used.
As explained above, the RDB is simple in structure and be easily understood. And thus the RDB has an advantage that constructions of a database, a creation of an application program and maintenance are easy. On the other hand, the RDB is involved in such a problem that the association among tables must be strictly defined for constructions. Further, as shown in part (b) of FIG. 22, there is a possibility that a large number of records free from data associated with items such as “degree”, “class” and “malt” of the “beverages” table exist. Thus, the RDB is also involved in such a problem that utility efficiency of storage areas of a computer is lowered and an access speed to a database is lowered.
FIG. 23 is a view showing an example of ODB data.
Usually, in the ODB, ones corresponding to tables and items of the RDB are referred to as “class” and “attribute”, respectively.
In the ODB of the name “classes” shown in FIG. 23, ones such as “maker”, “drink”, “alcohol”, “beer”, “sake”, “soft drink”, and “natural water”, which are coupled by solid lines, are “classes”. And ones such as “maker code”, “name” and “malt”, which are coupled with the associated upper hierarchy of classes by broken lines, respectively, are “attributes”. It is sufficient for each class to have inherent information. The upper hierarchy of attribute information is automatically succeeded to. For example, a definition of only the peculiar attributes to the “beer” such as “malt” and “country code” causes all of the “degree” attribute of the upper class “alcohol”, and “name” attribute, “age bracket” attribute, . . . and “small classification” attribute of the upper class “beverages” to be succeeded to as the attribute of the “beer” class. The ODB has the characteristic as mentioned above, and thus it is very free in the association among the classes. In view of such a characteristic, the ODB is being widely used in fields of the Internet and multimedia associations.
Hitherto to change over the RDB to the ODB, it is necessary to perform the following transition works with manual operation.                (1) analysis of RDB data        (2) study of a transition scheme to the ODB        (3) creation of program of extracting transition data from the RDB        (4) creation of data storage program to the ODB        
Thus, there is a need to manage with medium such as a paper various sorts of management information such as definition information of the necessary RDB and definition information of the ODB. Those works need a long time. And hitherto there is no transition-supporting tool for supporting those transition works. This is associated with a problem that errors due to manual operation are easy to happen.
As the database manipulation language for RDB, an SQL (Structured Query Language) is mainly used, and as the database manipulation language for ODB, an ODQL (Object oriented Database Query Language) is mainly used. To convert the application program of the RDB into the application program of the ODB, there is a need to perform a conversion work by the SQL and the ODQL which are quite different from each other. This requires a conversion work of a long term being equal to a term for newly developing the application program.